Bernhard Eversberg wrote:
Some metadata creators are inclined to follow no rules except their own, not disclosing what these are. But OK, we should not be pointing fingers at them but try very hard to make sense of everything they might come up with, creating a grand mashup (resisted to write hotchpotch.)
If that is so, and if metadata creators are not interested in getting the most out of our stuff either, why do we keep following extremely complex rules requiring innumerable elements? Dumb down RDA and MARC so we have only one elementfor keyword indexable text, and a few indispensable codes and dates. Wouldn’t that immensely ease the job of creating the mashup? After all, what more is Google doing, and who except us is saying that’s not good enough?
I think that each group sincerely believes its own standard to be better than anyone else’s. (I believe it!) So long as everyone holds onto such ideas, there can be no change and the result will be that a separate metadata record will forever be made and remade by each metadata community (or when taken to a reductio ad absurdum, even each library/bibliographical agency). This is the situation as it has always been, but before the WWW it was practically impossible to know about and share records with all of the other bibliographic agencies. Those difficulties have now been overcome. This situation becomes uncomfortable however, since earlier, while we honestly could not see the records produced by others, today we either have to pretend not to see them or willfully ignore them. This results in a situation that I don’t believe serves anybody very well.
The practice of cataloging is based on the principle of “consistency” which can turn cataloging into the most conservative of endeavors. By following the principle of consistency, catalogers ensure that the records they make today must work with the older records, some of them made 100 or more years ago. If you don’t keep this in mind, the result can be hiding the previous records or at least making those earlier records incomprehensible. Of course, lots of practices have changed tremendously, but the basic idea is for everything to work together. Can the principle of consistency be retained in an open, shared, cooperative environment? I think it can.
Perhaps I’m a dreamer, but since it seems as if the general public wants reliable metadata (ref. the Language Log discussion about the metadata in Google Books) I still think that it’s not too late, so long as catalogers are willing to adapt to some different practices. If we could simply get the rules pertaining to each separate bit of metadata, e.g. these page numbers follow the rules of the FAO of the UN, or by CERN, AACR2, Dewey, etc., it could go a long way for making the information more understandable.
I emphasize that this would be for librarians, who need this level of detail for their work of maintaining the collection, and not for users, who rarely need anything like this.